On Dec 6, 2013, at 8:20 PM, Robert Collins <[email protected]> wrote:

> On 7 December 2013 09:31, Liz Blanchard <[email protected]> wrote:
>> This list is great, thanks very much for taking the time to write this up! I 
>> think a big part of the User Experience design is to take a step back and 
>> understand the requirements from an end user's point of view…what would they 
>> want to accomplish by using this UI? This might influence the design in 
>> certain ways, so I've taken a cut at a set of user stories for the Icehouse 
>> timeframe based on these requirements that I hope will be useful during 
>> discussions.
>> 
>> Based on the OpenStack Personas[1], I think that Anna would be the main 
>> consumer of the TripleO UI, but please let me know if you think otherwise.
>> 
>> - As an infrastructure administrator, Anna needs to deploy or update a set 
>> of resources that will run OpenStack (This isn't a very specific use case, 
>> but more of the larger end goal of Anna coming into the UI.)
>> - As an infrastructure administrator, Anna expects that the management node 
>> for the deployment services is already up and running and the status of this 
>> node is shown in the UI.
>> - As an infrastructure administrator, Anna wants to be able to quickly see 
>> the set of unallocated nodes that she could use for her deployment of 
>> OpenStack. Ideally, she would not have to manually tell the system about 
>> these nodes. If she needs to manually register nodes for whatever reason, 
>> Anna would only want to have to define the essential data needed to register 
>> these nodes.
> 
> I want to challenge this one. There are two concerns conflated. A)
> seeing available resources for scaling up her cloud. B) minimising
> effort to enroll additional resources. B) is a no-brainer. For A)
> though, as phrased, we're talking about seeing a set of individual
> items: but actually, wouldn't aggregated capacity being more useful,
> with optional drill down - '400 cores, 2TB RAM, 1PB of disk'

Good point. I will update this to read that the user wants to see the available 
capacity and have the option to drill in further. [1]

> 
>> - As an infrastructure administrator, Anna needs to assign a role to each of 
>> the necessary nodes in her OpenStack deployment. The nodes could be either 
>> controller, compute, networking, or storage resources depending on the needs 
>> of this deployment.
> 
> Definitely not: she needs to deliver a running cloud. Manually saying
> 'machine X is a compute node' is confusing an implementation with a
> need. She needs to know that her cloud will have enough capacity to
> meet her users needs; she needs to know that it will be resilient
> against a wide set of failures (and this might be a dial with
> different clouds having different uptime guarantees); she may need to
> ensure that some specific hardware configuration is used for storage,
> as a performance optimisation. None of those needs imply assigning
> roles to machines.
> 
>> - As an infrastructure administrator, Anna wants to review the distribution 
>> of the nodes that she has assigned before kicking off the "Deploy" task.
> 
> If by distribution you mean the top level stats (15 control nodes, 200
> hypervisors, etc) - then I agree. If you mean 'node X will be a
> hypervisor' - I thoroughly disagree. What does that do for her?

We are in agreement, I'd expect the former. I've updated the use case to be 
more specific. [1] 

> 
>> - As an infrastructure administrator, Anna wants to monitor the deployment 
>> process of all of the nodes that she has assigned.
> 
> I don't think she wants to do that. I think she wants to be told if
> there is a problem that needs her intervention to solve - e.g. bad
> IPMI details for a node, or a node not responding when asked to boot
> via PXE.
> 
>> - As an infrastructure administrator, Anna needs to be able to troubleshoot 
>> any errors that may occur during the deployment of nodes process.
> 
> Definitely.
> 
>> - As an infrastructure administrator, Anna wants to monitor the availability 
>> and status of each node in her deployment.
> 
> Yes, with the caveat that I think instance is the key thing here for
> now; there is a lifecycle aspect where being able to say 'machine X is
> having persistent network issues' is very important, as a long term
> thing we should totally aim at that.
> 
>> - As an infrastructure administrator, Anna wants to be able to unallocate a 
>> node from a deployment.
> 
> Why? Whats her motivation. One plausible one for me is 'a machine
> needs to be serviced so Anna wants to remove it from the deployment to
> avoid causing user visible downtime.'  So lets say that: Anna needs to
> be able to take machines out of service so they can be maintained or
> disposed of.
> 
>> - As an infrastructure administrator, Anna wants to be able to view the 
>> history of nodes that have been in a deployment.
> 
> Why? This is super generic and could mean anything.
> 
>> - As an infrastructure administrator, Anna needs to be notified of any 
>> important changes to nodes that are in the OpenStack deployment. She does 
>> not want to be spammed with non-important notifications.
> 
> What sort of changes do you mean here?

Good question :) Definitely any failures. I was also thinking there could be 
certain thresholds on metrics like CPU that a user might want to be notified 
on. I'll add specific thoughts to the use case. [1] Thoughts?

> 
> ----
> 
> Thanks for putting this together, I love Personas as a way to make
> designs concrete and connected to user needs.

Thanks very much for all of the discussion!

Liz
[1] https://wiki.openstack.org/wiki/TripleO/Tuskar/IcehouseUserStories

> 
> -Rob
> 
> 
> -- 
> Robert Collins <[email protected]>
> Distinguished Technologist
> HP Converged Cloud
> 
> _______________________________________________
> OpenStack-dev mailing list
> [email protected]
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev


_______________________________________________
OpenStack-dev mailing list
[email protected]
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Reply via email to